A short course about fitting models with the scipy.optimize module

Fitting models and testing the match of the models to the measured data is a fundamental activity in many fields of science. This short (approximately 3-hour) course (available at: https://github.com/arokem/scipy-optimize) aims to teach participants to use the Scipy library’s optimize module to fit models to data (Jones et al. 2001). Using data from a psychology experiment (Rokem and Landau 2016) as an example, the course motivates the use of explicit mathematical models to explain and predict data and compares linear models and non-linear models. The core of the lesson focuses on fitting a curve with the curve_fit function. The course also introduces the idea of model comparison with cross-validation for evaluation and selection between non-nested non-linear models.